<![CDATA[Shopping an item in online store is a common activity happening to the community now. The rise of time makes someone chooses to shop online rather than having to travel to the store to get what they need. Reviews of each items in an online store can be useful to see how the buyer's previous feedback through a comment. The comments categorized as positive comments or negative comments. Therefore, to overcome the problem then used sentiment analysis reviews of items using Support Vector Machine and Query Expansion method. This research uses 400 data comments that is divided into two comment, that is positive and negative. The method used is Support Vector Macine polynomial kernel with degree two and Query Expansion. Query Expansion is used to expand a word that has synonyms that are not contained in the training data. The final test result yields an average of accuracy is 96,25% with parameter value of learning rate = 0,001, value of lambda = 0,1, value of complexity = 0,01 and maximum iteration is 50. Accuracy of Support Vector Machine and Query Expansion method is better than just using Support Vector Machine method which only gets 94,75% of accuracy.]]>
Copyrights © 2018
